Add complex64 support for tf.abs().

src/kernels/backend_cpu.ts

@@ -1107,13 +1107,21 @@ export class MathBackendCPU implements KernelBackend {
   }
 
   abs<T extends Tensor>(x: T): T {
-    this.assertNotComplex(x, 'abs');
-
     const resultValues = new Float32Array(x.size);
     const values = x.dataSync();
-    for (let i = 0; i < values.length; ++i) {
-      resultValues[i] = Math.abs(values[i]);
+
+    if (x.dtype === 'complex64') {
+      for (let i = 0; i < x.size; ++i) {
+        const real = values[i * 2];
+        const imag = values[i * 2 + 1];
+        resultValues[i] = Math.sqrt(real * real + imag * imag);
+      }
+    } else {
+      for (let i = 0; i < values.length; ++i) {
+        resultValues[i] = Math.abs(values[i]);
+      }
     }
+
     return Tensor.make(x.shape, {values: resultValues}) as T;
   }
 

src/kernels/backend_webgl.ts

@@ -49,6 +49,7 @@ import {BinaryOpComplexProgram} from './webgl/binaryop_complex_gpu';
 import * as binaryop_gpu from './webgl/binaryop_gpu';
 import {BinaryOpProgram} from './webgl/binaryop_gpu';
 import {ClipProgram} from './webgl/clip_gpu';
+import {ComplexAbsProgram} from './webgl/complex_abs_gpu';
 import {ConcatProgram} from './webgl/concat_gpu';
 import {Conv2DDerFilterProgram, Conv2DDerInputProgram} from './webgl/conv_backprop_gpu';
 import {DepthwiseConv2DDerFilterProgram, DepthwiseConv2DDerInputProgram} from './webgl/conv_backprop_gpu_depthwise';
@@ -1231,6 +1232,17 @@ export class MathBackendWebGL implements KernelBackend {
   }
 
   abs<T extends Tensor>(x: T): T {
+    if (x.dtype === 'complex64') {
+      const xData = this.texData.get(x.dataId);
+
+      const program = new ComplexAbsProgram(x.shape);
+      const inputs = [
+        this.makeComplexComponentTensorHandle(x, xData.complexTensors.real),
+        this.makeComplexComponentTensorHandle(x, xData.complexTensors.imag),
+      ];
+
+      return this.compileAndRun<Tensor>(program, inputs) as T;
+    }
     const program = new UnaryOpProgram(x.shape, unary_op.ABS);
     return this.compileAndRun(program, [x]) as T;
   }

src/kernels/webgl/complex_abs_gpu.ts

@@ -0,0 +1,37 @@
+/**
+ * @license
+ * Copyright 2018 Google LLC. All Rights Reserved.
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ * =============================================================================
+ */
+
+import {GPGPUProgram} from './gpgpu_math';
+
+export class ComplexAbsProgram implements GPGPUProgram {
+  variableNames = ['real', 'imag'];
+  userCode: string;
+  outputShape: number[];
+
+  constructor(shape: number[]) {
+    this.outputShape = shape;
+    this.userCode = `
+      void main() {
+        float real = getRealAtOutCoords();
+        float imag = getImagAtOutCoords();
+        vec2 v = vec2(real, imag);
+
+        setOutput(sqrt(dot(v, v)));
+      }
+    `;
+  }
+}

src/ops/unary_ops_test.ts

@@ -132,6 +132,42 @@ describeWithFlags('abs', ALL_ENVS, () => {
     expectArraysClose(result, [1, 2, 5, 3, 1, 4, 7, 8]);
   });
 
+  it('complex64 rank-1', () => {
+    const a = tf.complex([-2, -1, 0, 1, 2], [1, 2, 3, 0, -1]);
+    const result = tf.abs(a);
+    expectArraysClose(result, [
+      Math.sqrt(-2 * -2 + 1 * 1), Math.sqrt(-1 * -1 + 2 * 2),
+      Math.sqrt(0 * 0 + 3 * 3), Math.sqrt(1 * 1 + 0 * 0),
+      Math.sqrt(2 * 2 + -1 * -1)
+    ]);
+    expect(result.shape).toEqual([5]);
+  });
+
+  it('complex64 rank-2', () => {
+    const a = tf.complex([[-3, -2, -1], [0, 1, 2]], [[4, 1, 2], [3, 0, -1]]);
+    const result = tf.abs(a);
+    expectArraysClose(result, [
+      Math.sqrt(-3 * -3 + 4 * 4), Math.sqrt(-2 * -2 + 1 * 1),
+      Math.sqrt(-1 * -1 + 2 * 2), Math.sqrt(0 * 0 + 3 * 3),
+      Math.sqrt(1 * 1 + 0 * 0), Math.sqrt(2 * 2 + -1 * -1)
+    ]);
+    expect(result.shape).toEqual([2, 3]);
+  });
+
+  it('complex64 rank-3', () => {
+    const a = tf.complex(
+        [[[-3, -2], [-1, 0]], [[1, 2], [3, 4]]],
+        [[[4, 1], [2, 3]], [[0, -1], [-3, -4]]]);
+    const result = tf.abs(a);
+    expectArraysClose(result, [
+      Math.sqrt(-3 * -3 + 4 * 4), Math.sqrt(-2 * -2 + 1 * 1),
+      Math.sqrt(-1 * -1 + 2 * 2), Math.sqrt(0 * 0 + 3 * 3),
+      Math.sqrt(1 * 1 + 0 * 0), Math.sqrt(2 * 2 + -1 * -1),
+      Math.sqrt(3 * 3 + -3 * -3), Math.sqrt(4 * 4 + -4 * -4)
+    ]);
+    expect(result.shape).toEqual([2, 2, 2]);
+  });
+
   it('propagates NaNs', () => {
     const a = tf.tensor1d([1, -2, 0, 3, -0.1, NaN]);
     const result = tf.abs(a);